JimHokanson
commented
10 years ago Off the top of my head, these are the things that this repo does not have that the original repo did. Original repo: https://github.com/openworm/SegWorm/
1) Full video parsing. This has been started. See seg_worm.segWormFrames 2) feature info filtering - somewhere I've seen code that is meant to select features from videos that fufill certain requirements - recording date, strain, on or off food, etc. 3) Displaying things - This is a huge hole that is missing. There are tons of different plotting routines to summarize things that have not been reproduced.
I think that's it. Points 2 & 3 are relatively straightforward, in terms of complexity, so I figured I didn't need to do them in Matlab.
Things that would be valuable to convert:
- some mechanism for managing feature files - ftp downloading, filtering by sql queries
- displays that allow replication of published results
There are probably other displays to make but I think these should be driven by our desire to understand the statistical differences between things, not to necessarily replicate code
Hi @JimHokanson,
As part of achieving this milestone of full conversion of this repository to the movement_validation repository, I was wondering if you could inventory those things that would be valuable to convert, even if they are not totally completed.
We've converted all the features code, and I'm working on the stats code, however, I'm believe you did some other work, on the computer vision piece, and perhaps on the GUI or perhaps other things?
And even things that you didn't work on might be worth bringing over if the eventual goal is a full port. As discussed the top priority is to finish statistics, but perhaps it would be nice to at least have a list of what else.
Thanks!